You probably think of virtual consults for skin issues as a simple win. Members get a quick diagnosis, a prescription in 24 hours, and you save a few bucks compared to an in-person specialist visit. But if you look under the hood of your benefits systems, a different picture emerges. Some telederm platforms quietly degrade your plan data, route prescriptions around your PBM, and create liability risks nobody flagged during the RFP.
The clinical care is often fine. The failure is administrative: these platforms connect to your benefit systems, or fail to, in ways that turn a promising benefit into a data black hole.
Low-resolution claims data
A standard doctor visit produces clean, structured data: a CPT code, an ICD-10 diagnosis, a provider NPI. That data feeds your analytics, your stop-loss reporting, your population health models. It’s how you know what’s happening across your plan.
A telederm visit looks different. A member takes three photos of a rash. The platform’s AI, or a board-certified dermatologist, reviews the images and identifies a suspicious mole, recommending an urgent biopsy. Your claims system records a single code for a brief virtual check-in, CPT 98016, and nothing more.
The platform knows the member is at risk, but your analytics, your care management team, and your stop-loss carrier never see it. The critical data point, “this patient may have melanoma,” never gets recorded in the systems that matter. That is the pixel-versus-code conflict. You are paying for high-quality image analysis while your admin system captures a low-resolution data point.
Prescriptions that skip the PBM
The risk is concentrated in direct-to-consumer-style platforms, which do more than diagnose. They can route a prescription to their own or an affiliated cash-pay pharmacy, where branded or compounded topicals are billed outside your PBM contract.
Take tretinoin. Your PBM may have negotiated a steep discount on the generic. A platform that dispenses its own branded or compounded formulation bypasses that price, the member pays out of pocket or the cost hides in a vendor fee, and the claim never shows up in your pharmacy data. A 2022 study in the Annals of Internal Medicine found that direct-to-consumer telehealth platforms charged a median of 2.3 times the national average drug acquisition cost for common generics. That spread is where the vendor makes its margin, which is why the arrangement persists.
This is a shadow pharmacy network. It undermines your PBM contract and inflates costs in ways your standard reports can’t detect. It rarely comes up in vendor due diligence.
Cross-state licensure gaps
A dermatologist licensed in Texas treats one of your members who is vacationing in Florida. The platform auto-assigns the doctor, and the claim is filed with the Texas NPI and a telehealth place-of-service code. Florida law lets an out-of-state physician treat Florida patients, but only after the physician completes an out-of-state telehealth registration with the Florida Department of Health (section 456.47, Florida Statutes) or holds a Florida license. That registration is rarely checked.
Your credentialing system rarely integrates with the telederm vendor’s provider panel. If that claim is ever audited, during an ERISA lawsuit or a subrogation dispute, you have no way to prove after the fact that the rendering doctor was authorized to treat the member in Florida, unless the vendor captured the license or registration at the time of the consult. That is a benefit administration liability that catches employers off guard.
State-by-state licensure rules
The licensure risk is real, but it is checkable. Two mechanisms matter. The Interstate Medical Licensure Compact gives eligible physicians an expedited path to licenses in member states, and it reached 44 states by mid-2026, but it is not a multistate license. A compact physician still holds a separate license in each state where a member sits. Separately, some states run telehealth registrations. Florida’s registration requires a designated registered agent and liability coverage, with no full Florida license needed.
Neither mechanism is automatic, which is why it belongs in the RFP. The requirement is that the vendor captures, at the time of each consult, that the rendering clinician held an active license or registration in the member’s state. Being licensed somewhere is not enough. That turns an after-the-fact blind spot into a contract term you can audit.
Low-severity coding and stop-loss risk
Your actuaries rely on that same claims feed. Telederm platforms generate episodic claims with low diagnostic specificity. A member with severe psoriasis uses the app for quick flares, and the platform records “mild plaque psoriasis” instead of the more complex “psoriasis with arthropathy.” Later, that member needs a biologic. A 2025 study in JAMA Dermatology put the average annual net cost of first-line psoriasis biologics at $47,125 in 2021.
Your stop-loss carrier reviews that history and sees a pattern of cheap virtual visits with low-severity codes. It may challenge the large claim or raise your rates because the risk data looks inconsistent. The clinical care may have been fine. The data pollution is what makes your plan look healthier than it is.
RFP requirements for telederm vendors
Stop evaluating teledermatology on member satisfaction scores alone. Evaluate it on data interoperability, and put these requirements in your next vendor RFP:
- Require a structured data feed. Don’t accept a basic EDI claim file. Ask for an API that delivers image metadata, provider license or registration in the member’s state at the time of consult, referral paths (e.g., “required in-person biopsy within 7 days”), and NDC codes for any drug dispensed directly by the vendor.
- Force pharmacy integration. Do not let the vendor use its own pharmacy. Route all prescriptions through your primary PBM. This closes the shadow leak and lets you audit drug costs properly.
- Audit diagnosis fidelity annually. Pull a sample of members with high-cost claims (biologics, Mohs surgery) who also used the telederm benefit. Check whether the platform’s initial diagnosis matches the specialist’s final diagnosis. If the mismatch rate exceeds 5%, your stop-loss credibility is eroding.
Virtual dermatology is often a clinical win. From a benefits systems perspective, though, it can be a data pollution event. Demand fidelity of data, fidelity of prescribing, and fidelity of reporting. Otherwise you are buying a feature that looks good on paper but quietly undermines your plan.